Development of a fault detection algorithm for Photovoltaic Systems

The use of an algorithm based on data from I-V curves, a simple and cost-effective method for fault detection and identification in Photovoltaic Systems (PVS), is presented. When determining whether or not to invest in a PVS, life expectancy and reliability are critical considerations. In this paper, the development of an I-V curve-based algorithm for fault detection and identification in PVS is presented. The method calculates the single diode model that describes the Photovoltaic cell in use, for the irradiance and temperature of a certain location. After that, a threshold monitoring approach identifies the presence and the nature of a fault. Measurements were performed to certify the ability of the algorithm to detect both the normal operation at maximum power point and the ability to detect and identify errors introduced during the operation of the experiment. The algorithm can identify open-circuit, short-circuit and mismatch faults. The results are promising, implying that the method could be applied in PVS.